No abstract

The storage of voltage and current signals over a period of time generates a large memory expense. Therefore, signal compression techniques became important in this context. This paper presents an implementation in Field Gate Programmable Array (FPGA) of an algorithm of sparse representation using redundant dictionaries, applied to the compression of electrical signals, from power systems. The representation will be based on a dictionary constituted by elements of the Fourier and Wavelet basis, that are capable to represent the stationary and transient components of the electrical signals. The results will be analyzed due to two parameters: the quality of the compressed signal, in terms of its correlation coefficient related to the original signal; and the number of elements in te representation, that is related with the compression ratio. The feasibility of the implementation in real time will be evaluated in terms of the consumed FPGA resources and the necessary frequency of operation.

Harmonic impedance is an important parameter for electrical power system analyses. This parameter can be used in some important application such as island and resonance detection. The methods used for impedance estimation can be separated in two categories: invasive and noninvasive methods. The noninvasive methods do not cause any disturbance in the grid, but the results are less accurate than the invasive methods. However, in both methods, when there is frequency deviation, the results are less accurate, due to the spectral leakage caused by asynchronous sample rate. Then, this work presents a technique to improve the results of the impedance estimation using Lagrange Interpolation to synchronize the sampling and to avoid the spectral leakage.

The objective of this paper was to examine the dynamic impedance estimation of electrical systems from online measurements. The paper makes several considerations and highlights the challenges to obtain a precise estimation. Transducer equalization and harmonic synchrophasor estimation (HSpE) are reviewed and discussed. The use of online and adaptive equalization for transducers proves to be a viable solution for improving voltage transducer’s (VT’s) and current transducer’s (CT’s) frequency response. Additionally, the use of oversampling algorithms can mitigate the effects of noise in the HSpE. Furthermore, methods for harmonic impedance estimation are discussed. The independent component analysis ICA-based dynamic impedance estimation is proposed and results presented, which yields excellent agreement. Finally, harmonic modeling and simulation of injected harmonic currents are used to observe resonances through the amplification and attenuations and, consequently, the opportunity to confirm the system self and transfer impedances of a test system. Dynamic impedance estimation will continue to be a great challenge for the power systems engineer as the system complexity increases with the massive insertion of power electronic inverters and the associated required filtering. Real-time signal processing will be an effective tool to determine the dynamic self or transfer impedance.

No abstract